Gmail data access should be purpose-limited: retrieve the minimum fields needed for a declared feature, preserve provider IDs, and delete cached content when that purpose ends.
An inbox summary, contact timeline, attachment search, and compliance archive need different data. Write the exact fields, retention period, and user-visible value first; then choose the narrowest Gmail scopes and formats that support it.
List candidate message IDs, fetch metadata or bodies selectively, parse MIME defensively, and stream large attachments rather than loading them into worker memory. Keep raw provider identifiers for reconciliation.
Several Gmail read and modify scopes are restricted. Production apps can require OAuth verification, and storing or transmitting restricted-scope data on servers can require a security assessment. A unified API cannot remove those obligations.
Encrypt necessary cached content, separate tenants, redact logs, and provide account disconnect and deletion workflows that remove tokens and derived mailbox data according to the product policy.
Consent is not a blanket retention license. Keep data only for the disclosed purpose and period, and follow Google's user-data policies plus the customer's legal obligations.
No. The customer-facing Google application and its use of scopes must satisfy Google's consent, verification, and security requirements.
Not yet. Gmail, Microsoft Graph, and IMAP/SMTP connectors are planned. This guide documents the intended architecture and the provider requirements a production release must satisfy.
DewEngine is in development. Real use cases decide what ships first.